RARE-57. PEDIATRIC CHORDOMA: WHOLE EXOME SEQUENCING OF 11 PEDIATRIC CHORDOMA SAMPLES

Chordoma is a rare tumor and while SMARCB1 alterations have been observed in poorly differentiated chordomas, conventional chordomas are not well understood. We interrogated nuclear and mitochondrial genomes of 11 chordoma samples from 7 children. Frozen tumor tissue DNA was extracted and whole exome libraries generated using Agilent SureSelect Human All Exon V6 kit plus mtDNA genome capture kit. Libraries were sequenced using Illumina Nextseq 500. MuTect2, VarDict and LUBA variant callers were used with allele frequency cutoff 2%. Potential germline variants were filtered bioinformatically. In total, 656±74 high-confidence somatic variants, including 368±43 nonsynonymous variants per sample were detected. Of 2,607 combined unique nonsynonymous variants, 95% were missense. Remaining high impact variants were frameshift (37%), stop gain (39%), splice acceptor/donor (22%), start and stop loss (2%). Of the unique nonsynonymous variants, 137 fall within Cosmic Cancer Census Genes, including high impact variants in SETD2, MLLT4. No previously reported TBXT, CDKN2A, PI3K, LYST mutations identified. Tumor Mutation Burden/Megabase was 10±1. The mitochondrial analysis revealed heteroplasmic m.11727C>T MT-ND4 missense variants in three tumors resected at different time points from the same patient, and another heteroplasmic m.1023C>T rRNA mutation from the primary and recurrent tumors of another patient. Intriguingly, two Children’s Brain Tumor Tissue Consortium patients with chordoma had identical heteroplasmic m.10971G>A MT-ND4 nonsense mutations. Pediatric chordomas appear to lack somatic nuclear mutations. Observing recurrent mitochondrial mutations across multiple tumors from the same and/or different patients is striking, suggesting they may be implicated in tumorigenesis and be potential diagnostic markers.

Bioinformatics Pipeline for Human Papillomavirus Short Read Genomic Sequences Classification Using Support Vector Machine

We recently developed a test based on the Agilent SureSelect target enrichment system capturing genomic fragments from 191 human papillomaviruses (HPV) types for Illumina sequencing. This enriched whole genome sequencing (eWGS) assay provides an approach to identify all HPV types in a sample. Here we present a machine learning algorithm that calls HPV types based on the eWGS output. The algorithm based on the support vector machine (SVM) technique was trained on eWGS data from 122 control samples with known HPV types. The new algorithm demonstrated good performance in HPV type detection for designed samples with 25 or greater HPV plasmid copies per sample. We compared the results of HPV typing made by the new algorithm for 261 residual epidemiologic samples with the results of the typing delivered by the standard HPV Linear Array (LA). The agreement between methods (97.4%) was substantial (kappa = 0.783). However, the new algorithm identified additionally 428 instances of HPV types not detectable by the LA assay by design. Overall, we have demonstrated that the bioinformatics pipeline is an accurate tool for calling HPV types by analyzing data generated by eWGS processing of DNA fragments extracted from control and epidemiological samples.

Optimizing nucleic acid extraction and transcriptome evaluation from low-input, fixed clinical samples

Tumor biopsies are commonly formalin-fixed and paraffin-embedded (FFPE) for long-term and efficient storage. However, FFPE preservation can greatly compromise the quality of samples, the extraction of nucleic acids, and feasibility of downstream studies, including RNA sequencing. These challenges are especially evident in the studies of clinical trial samples, where the sizes of biopsies often limit the amount of material available for study. Here, we evaluate two nucleic acid extraction kits (Covaris truXTRAC FFPE tNA Plus, QIAGEN miRNeasy FFPE) and three hybridized-capture-based RNA sequencing library preparations (Agilent SureSelect XT RNA Direct, Agilent SureSelect XT HS, and Illumina TruSeq RNA Exome) to evaluate the impact of these sample processing steps on transcriptome evaluation in a melanoma biopsy procured in a clinical setting and preserved by FFPE. While there exist many options for extraction and library preparation that are appropriate for RNA sequencing from FFPE samples, we observed that combinations of experimental approaches may have subtle impacts on downstream analysis, including gene expression quantification and fusion detection.

PL3.3 Second interim and first molecular analysis of the EORTC randomized phase III intergroup CATNON trial on concurrent and adjuvant temozolomide in anaplastic glioma without 1p/19q codeletion

BACKGROUND The 1st interim analysis of the CATNON trial showed benefit from adjuvant (adj) temozolomide (TMZ) on overall survival (OS) but remained inconclusive about concurrent (conc) TMZ. A 2nd interim analysis was planned after 356 events. MATERIAL AND METHODS The 2x2 factorial design phase III CATNON trial randomized 751 adult patients with newly diagnosed non-codeleted anaplastic glioma to either 59.4 Gy radiotherapy (RT) alone; the same RT with concTMZ; the same RT and 12 cycles of adjTMZ or the same RT with both concTMZ and adjTMZ (doi: 10.1016/S0140-6736(17)31442-3). MGMT promoter methylation (MGMTmeth) status was re-assessed with the Infinium Methylation EPIC Beadchip using the MGMT_STP27 model. Isocitrate dehydrogenase 1 and 2 (IDH) mutation (mt) status was assessed with glioma targeted Agilent SureSelect baits sequence using an Illumina HiSeq2500 Rapid PE100. RESULTS With a median follow-up of 56 months and 356 events, the hazard ratio (HR) for OS adjusted for stratification factors after concTMZ was 0.968 (99.1% CI 0.73, 1.28). 5-year OS was 50.2% with and 52.7% without concTMZ (95% CI [44.4, 55.7] and [46.9, 58.1]). An IDHmt was found in 335 of 480 assessed cases (70%). Median OS was 19 mo (95% CI 16.3, 22.3) in IDHwt tumors and 116 mo (95% CI 82.0, 116.6) in IDHmt tumors. The interaction test based on IDH status was significant (p=0.016) in the univariate HR analysis for OS after concTMZ (IDHwt, n=145, events=120, HR = 1.27, 95% CI 0.89, 1.82, p=0.19; IDHmt, n=335, events=92, HR= 0.67, 95% CI 0.44, 1.03, p=0.06). IDHmt was predictive of benefit from adjTMZ (IDHmt HR: 0.41, 95% CI 0.27, 0.64; IDHwt: HR 1.05, 95% CI 0.73, 1.52; interaction test p = 0.001). In IDHmt patients that received adjTMZ, the HR for OS after concTMZ was 0.71 (95% CI 0.35, 1.42, p=0.32). MGMTmeth was found in 288 of 410 assessed cases (70%), interaction test for concTMZ (p = 0.092) and adjTMZ (p = 0.166) did not reach statistical significance. CONCLUSION In the entire study cohort, concTMZ did not increase OS. However, in IDHmt tumors a trend towards benefit of concTMZ is present. AdjTMZ increased OS in IDHmt but not in IDHwt tumors. Further analyses and follow-up will allow full assessment of efficacy in the molecular subgroups.

Second interim and first molecular analysis of the EORTC randomized phase III intergroup CATNON trial on concurrent and adjuvant temozolomide in anaplastic glioma without 1p/19q codeletion.

2000 Background: The 1st interim analysis of the CATNON trial showed benefit from adjuvant (adj) temozolomide (TMZ) on overall survival (OS) but remained inconclusive about concurrent (conc) TMZ. A 2nd interim analysis was planned after 356 events. Methods: The 2x2 factorial design phase III CATNON trial randomized 751 adult patients with newly diagnosed non-codeleted anaplastic glioma to either 59.4 Gy radiotherapy (RT) alone; the same RT with concTMZ; the same RT and 12 cycles of adjTMZ or the same RT with both concTMZ and adjTMZ ( doi: 10.1016/S0140-6736(17)31442-3). MGMT promoter methylation ( MGMTmeth) status was re-assessed with the Infinium Methylation EPIC Beadchip using the MGMT_STP27 model. Isocitrate dehydrogenase 1 and 2 ( IDH) mutation (mt) status was assessed with glioma targeted Agilent SureSelect baits sequence using an Illumina HiSeq2500 Rapid PE100. Results: With a median follow-up of 56 months and 356 events, the hazard ratio (HR) for OS adjusted for stratification factors after concTMZ was 0.968 (99.1% CI 0.73, 1.28). 5-year OS was 50.2% with and 52.7% without concTMZ (95% CI [44.4, 55.7] and [46.9, 58.1]). An IDHmt was found in 335 of 480 assessed cases (70%). Median OS was 19 mo (95% CI 16.3, 22.3) in IDHwt tumors and 116 mo (95% CI 82.0, 116.6) in IDHmt tumors. HR for OS after concTMZ in patients with known IDH status. Clinical trial information: NCT00626990. IDHmt was predictive of benefit from adjTMZ ( IDHmt HR: 0.41, 95% CI 0.27, 0.64; IDHwt: HR 1.05, 95% CI 0.73, 1.52; interaction test p = 0.001). In IDHmt patients that received adjTMZ, the HR for OS after concTMZ was 0.71 (95% CI 0.35, 1.42, p=0.32). MGMTmeth was found in 288 of 410 assessed cases (70%), interaction test for concTMZ (p = 0.092) and adjTMZ (p = 0.166) did not reach statistical significance. Conclusions: In the entire study cohort, concTMZ did not increase OS. However, in IDHmt tumors a trend towards benefit of concTMZ is present. AdjTMZ increased OS in IDHmt but not in IDHwt tumors. The ongoing molecular analyses and further follow-up will allow full assessment of efficacy in the molecular subgroups.[Table: see text]

Clonal Trajectories in Persons with Acute Myeloid Leukemia and Normal Cytogenetics Relapsing after Intensive Chemotherapy and after Allotransplants. Does Pretransplant Conditioning Contribute to Mutation Topography?

Background One-half of persons with acute myeloid leukaemia (AML) have normal cytogenetics at diagnosis and a good prognosis. However, some relapse. Whether the clonal trajectory in these persons is like that of persons with unfavorable cytogenetics at diagnosis who relapse is unknown. Aims Interrogate the mutation topography and clone trajectory of adults with AML and normal cytogenetics at diagnosis achieving complete remission and relapsing after intensive chemotherapy (N=17), an allotransplant (N=4) or both (N=2). Methods We performed whole exome sequencing (WES) on diagnosis (N=23), remission (N=23) and relapse samples (N=25) from 23 subjects with AML and normal cytogenetics at diagnosis who relapsed after intensive chemotherapy with (N=6) or without an allotransplant (N=17) and from 6 normal transplant donors. For filtering germline polymorphisms corresponding remission samples with corresponding donor samples were controls. Sequencing libraries were generated using Agilent SureSelect Human All Exon® kit V5(Agilent Technologies, Santa Clara, CA, USA). Mean sequencing depths were 200X (range, 145-265X) for the relapse samples and 138X (range, 106-173X) for the remission and donor samples. Median coverage was 99.9% (range, 77.6-99.9 %). NPM1, FLT3-ITD and CEBPA were also tested by standard techniques. Results We detected 1317 non-synonymous somatic variants in 988 genes including 569 variants in 483 genes at diagnosis and 748 in 665 genes at relapse. 160 of these were concordant. Median numbers of mutations were both 23 (ranges,7-113 and 2-146). Most mutations were mis-sense including 84% (95% confidence interval [CI], 81, 87%)at diagnosis and 83% (80, 86%)at relapse. C>A/G>T transversions were the most common mutations at diagnosis and relapse. Frequency at relapse (52% [36, 38%]) was significantly higher than at diagnosis,37% (80, 86%; P=0.0405). This was also so for C>T/G>A transitions (24% [23, 25%] vs 21% [20, 22%]; P=0.074). Mutations in FLT3-ITD, NPM1, CEBPA, DNMT3A, IDH2, WT1 and GATA2 were frequent and typically concordant in diagnostic and relapse samples. However variable allele frequencies (VAFs) were significantly higher in relapse compared with diagnosis samples (median 41% [range, 4-86%] vs. 9% [range, 2-65%]; P<0.001). We also identified 23 novel mutated genes such as TUBA3D, MED16, PARP4 and ZADH2. There was also a significant increase in copy number variants (CNV) between diagnosis(N=17) and relapse (N=42) samples of which only 4 were shared. Mutation trajectories from diagnosis to remission to relapse were classified as: (1) similar spectrum (N=6); (2) evolution from a dominant clone at diagnosis (N=13); and (3) evolution from a minor clone at diagnosis (N=4). Conclusions Our data illustrate the mutation topography and clonal trajectories of adults with normal cytogenetics AML from diagnosis to relapse, identify 23 new mutated genes and indicate mutation-related relapse patterns. We suggest mutation(s) possibly caused by pretransplant conditioning contribute to posttransplant relapse. Figure. Figure. Disclosures No relevant conflicts of interest to declare.

Targeted enrichment outperforms other enrichment techniques and enables more multi-species RNA-Seq analyses

Enrichment methodologies enable analysis of minor members in multi-species transcriptomic analyses. We compared standard enrichment of bacterial and eukaryotic mRNA to targeted enrichment with Agilent SureSelect (AgSS) capture for Brugia malayi, Aspergillus fumigatus, and the Wolbachia endosymbiont of B. malayi (wBm). Without introducing significant systematic bias, the AgSS quantitatively enriched samples, resulting in more reads mapping to the target organism. The AgSS-enriched libraries consistently had a positive linear correlation with its unenriched counterpart (r2=0.559-0.867). Up to a 2,242-fold enrichment of RNA from the target organism was obtained following a power law (r2=0.90), with the greatest fold enrichment achieved in samples with the largest ratio difference between the major and minor members. While using a single total library for prokaryote and eukaryote in a single sample could be beneficial for samples where RNA is limiting, we observed a decrease in reads mapping to protein coding genes and an increase of multi-mapping reads to rRNAs in AgSS enrichments from eukaryotic total RNA libraries as opposed to eukaryotic poly(A)-enriched libraries. Our results support a recommendation of using Agilent SureSelect targeted enrichment on poly(A)-enriched libraries for eukaryotic captures and total RNA libraries for prokaryotic captures to increase the robustness of multi-species transcriptomic studies.

A modified sequence capture approach allowing standard and methylation analyses of the same enriched genomic DNA sample

BackgroundBread wheat has a large complex genome that makes whole genome resequencing costly. Therefore, genome complexity reduction techniques such as sequence capture make re-sequencing cost effective. With a high-quality draft wheat genome now available it is possible to design capture probe sets and to use them to accurately genotype and anchor SNPs to the genome. Furthermore, in addition to genetic variation, epigenetic variation provides a source of natural variation contributing to changes in gene expression and phenotype that can be profiled at the base pair level using sequence capture coupled with bisulphite treatment. Here, we present a new 12 Mbp wheat capture probe set, that allows both the profiling of genotype and methylation from the same DNA sample. Furthermore, we present a method, based on Agilent SureSelect Methyl-Seq, that will use a single capture assay as a starting point to allow both DNA sequencing and methyl-seq.ResultsOur method uses a single capture assay that is sequentially split and used for both DNA sequencing and methyl-seq. The resultant genotype and epi-type data is highly comparable in terms of coverage and SNP/methylation site identification to that generated from separate captures for DNA sequencing and methyl-seq. Furthermore, by defining SNP frequencies in a diverse landrace from the Watkins collection we highlight the importance of having genotype data to prevent false positive methylation calls. Finally, we present the design of a new 12 Mbp wheat capture and demonstrate its successful application to re-sequence wheat.ConclusionWe present a cost-effective method for performing both DNA sequencing and methyl-seq from a single capture reaction thus reducing reagent costs, sample preparation time and DNA requirements for these complementary analyses.

Complicated Staphylococcus aureus Bacteremia (SAB) Is Associated with Genetic Variation in GLS2

Background SAB is a serious, common infection. We used whole exome sequencing (WES) to examine the cumulative effect of coding variants in each gene on risk of complicated SAB in a discovery set of patients, and then evaluated the nominally significant genes in a replication set of patients using custom-capture sequencing. Methods The discovery set comprised 84 complicated SAB cases (endocarditis or bone/joint infection) frequency-matched by age (in deciles), sex, and bacterial clonal complex (CC5/30, CC8) to 84 uncomplicated SAB controls. All were white inpatients at Duke University. WES utilized Agilent SureSelect 72Mb capture kits, followed by sequencing on an Illumina HiSeq2000, alignment and base calling with a standard pipeline. The SKAT-O and EPACTS packages were used for gene-based association tests and logistic regression models with Firth bias correction, respectively. Both controlled for age, sex, and clonal complex as covariates. The replication set of 122 complicated SAB cases and 118 uncomplicated SAB controls was frequency matched by age, sex, and clonal complex. All were white Europeans collected by the Statens Serum Institute. An Agilent SureSelect 2Mb capture array captured genic sequence for 342 genes nominally associated with complicated SAB in discovery (SKAT-O P &lt; 0.035). Sequencing and data analysis proceeded as for WES. A Bonferroni-corrected gene-based test P-value of 1.5×10–4 determined significance in the replication set. Results One gene, GLS2, was significantly associated with complicated SAB in the replication set (P = 1.2 x 10–4). The strongest single-variant association in all 342 genes was rs2657878 in GLS2 (p = 5×10−4). This variant is strongly correlated with a missense variant (rs2657879, p = 4.4x10-3) in which the minor allele (associated here with complicated SAB) has previously been shown to reduce circulating glutamine levels. Conclusion Comprehensive examination of the coding sequence for association with complicated SAB in a two-stage discovery/replication design identified a novel candidate gene. GLS2 is an interesting candidate for complicated SAB due to its role in regulating glutamine production, a key factor in activation of T-cell production. Disclosures V. Fowler Jr., Pfizer, Novartis, Galderma, Novadigm, Durata, Debiopharm, Genentech, Achaogen, Affinium, Medicines Co., Cerexa, Tetraphase, Trius, MedImmune, Bayer, Theravance, Cubist, Basilea, Affinergy, Janssen, xBiotech, Contrafect: Consultant, Consulting fee; NIH, Basilea, MedImmune, Cerexa/Forest/Actavis/Allergan, Pfizer, Advanced Liquid Logics, Theravance, Novartis, Cubist/Merck; Medical Biosurfaces; Locus; Affinergy; Contrafect; Karius: Grant Investigator, Research grant; Green Cross, Cubist, Cerexa, Durata, Theravance; Debiopharm: Consultant, Consulting fee; UpToDate: author on several chapters, Royalties

Single Cell Whole Exome Sequencing in an Index Case of Amp1q21 Multiple Myeloma to Define Intraclonal Variation

Introduction Multiple myeloma (MM) is a largely incurable plasma cell malignancy characterised by marked genomic heterogeneity, in which chromosome 1q21 amplification (amp1q21) associates with poor prognosis. Genomic analysis using next generation sequencing has identified recurrent mutations, but no universal acquired somatic mutation(s) have emerged in MM, suggesting that understanding pathways of survival will require analysis of individual tumours in distinct disease subsets. To compound complexity of the problem, intraclonal variation (ICV), known as a major driver mechanism in cancer plasticity, in which clonal competitor cells undergo selection during disease evolution and progression by Darwinian principles, will need to be fully mapped at the genome level. Identifying the true level of ICV in a tumour will thus require analysis at the level of whole exome sequencing (WES) in single cells (SCs). In this study, we sought to establish WES methodology able to identify ICV in SCs in an index case of amp1q21 MM. Methods Cell selection and sequencing CD138+ tumour cells and CD3+ T-cells were isolated from a presentation case of amp1q21 MM as bulk populations to high purity (>97%). Single MM cells and normal T cells were individually isolated and used for single cell (SC) whole exome sequencing (WES). Whole genome amplification (WGA) was performed by multiple displacement amplification (Qiagen REPLI-g Mini kit), and exome capture was performed using Agilent SureSelect. Libraries were then 90 bp paired end sequenced on an Illumina HiSeq2000 (BGI, China). Data analysis Data was produced for bulk (1000 cells) MM and bulk germline T cells, twenty MM SCs and five T cell SCs. Raw data was aligned to hg19 reference sequence using NovoAlignMPI (v3.02.03). Variant calling was performed using SAMtools (v1.2.1) and VarScan (v2.3.6) and variants were annotated using ANNOVAR. High confidence variants were called in the bulk tumour WES by pairwise comparison with bulk germline WES. Variant lists were also cross-searched against various variant databases (CG46, 1000 genomes, dbSNP, esp650 and in-house database) in order to exclude variants that occur in the general population. Multiple quality control measures were employed to reduce the number of false positive calls. Results and Discussion Data and bioinformatics pipelines are of a high quality SC WES generated raw data reads that were similar to bulk WES of 1000 cells, with comparable mapping to Agilent SureSelect target exome (69-76% SC vs. 70% bulk) and mean fold coverage (56.8-59.1x vs. 59.7x bulk). On average, 82% of the exome was covered sufficiently for somatic variant (SV) calling (often considered as ≥ 5x), which was higher than seminal published SC WES studies (70-80%) (Hou et al., Cell, 2012; Xu et al., Cell, 2012). We identified 33 potentially deleterious SVs in the bulk tumour exome with high confidence bioinformatics, 21 of which were also identified in one or more SC exomes. The variants identified include suspected deleterious mutations in genes involved in MAPK pathway, plasma cell differentiation, and those with known roles in B cell malignancies. To confirm SV calls, randomly selected variants were validated by conventional Sanger sequencing, and of 15/15 variants in the bulk WES and of 55/55 variants in SCs, to obtain 100% concordance. Intraclonal variation in MM Significantly, ICV was apparent from the SC exome variant data. Total variant counts varied considerably among SCs and most variant positions had at least several cells where no evidence of the variant existed. Bulk WES lacks crucial information We identified an additional 23 variants that were present in 2+ SC exomes, but absent in the bulk MM tumour exomes. Of these, 30% (7 variants) were examined for validation, and were amplifiable in at least one cell to deliver 100% concordance with variant calls. These variants are of significant interest as they reveal a marked occurrence of subclonal mutations in the MM tumour population that are not identified by bulk exome sequencing. They indicate that the mutational status of the MM genome may be substantially underestimated by analysis at the bulk tumour population level. Conclusion In this work we establish the feasibility of SC WES as a method for defining intraclonal genetic variation in MM. Disclosures No relevant conflicts of interest to declare.